Even when those little gods of wine, the yeast, have done their fermentation job, they still do good work on wine, adding flavour, aroma, body, structure texture, weight, complexity, even petillance (that’s a faint sparkle, not a petulant little wine), and both protein and tartrate stability.

It’s important to discuss the right lees. Wine should always be taken of the gross lees (from the French for heavy), which in wine terms are gross (from the English for gross).  It is only the fine lees that are potentially good for wine.  

sur lie means ‘on the lees’

Essentially lees comprise dead yeast cells, and maybe some bits of grape skin and pulp that settle to the bottom of the vat at the end of fermentation.

This technique of leaving wine on fine lees mostly applies to white wines, but increasingly reds too are left on the fine lees for several months. The classic white wine example is Muscadet ‘sur lie’, from the Loire valley in France.  ‘Sur lie’ means simply ‘on the lees’, So this style should have a bit more weight, substance and character than a straight Muscadet.

Lees ageing is taken to its apogee in traditional method bubblies, such as Champagne, Cava, Franciacorta from Italy and Cap Classique from South Africa. 

Yeast autolysis (auto-breakdown of yeast cells by their own enzymes), as it’s called when used with bubblies, confers rich flavours of biscuit, brioche, white nuts such as almonds and macadamias, toast, even marmite notes which add depth, breadth and savoury character to the wine. Flavour crudely depends on the length of time the wine is in contact with the lees. So a Cava that has had the minimum nine months on lees will taste more of the original fruit than vintage Champagne, with at least three years on lees.  Autolysis is also argued to give fineness and persistence to bubbles.  And some argue that a greater integration of bubbles occurs with longer lees ageing.

stirring the lees lessens the risk of reductive conditions

Stirring the lees (battonage) in the barrel or vat is done to increase the effects of lees contact. This obviously cannot be done with traditional method bubbly, although some producers practise the art of poignettage where bubbly bottles are shaken to mix the yeast cells into the wine again, for example Gosset Champagne and Recaredo Cava. Stirring also minimises the risk of reductive flavours (think stink bombs and rotten eggs) occurring, which can develop under a thick layer of yeast cells. But slightly reductive conditions are good to help preserve primary fruit character, so it is a delicate balance to get the best from fine lees.

Mannoproteins are released naturally during lees contact. It is these that create a creamy, silky smooth mouthfeel, and a richer texture to the body of the wine. They are also important for wine stability.

A good, if expensive, comparison would be to choose a non-vintage and a vintage Champagne from the same producer.

Oxygen, friend and foe: can’t live without it, eventually die with it. Managing highly reactive oxygen gas during winemaking and bottling is like walking the tightrope. At one end is reduction, at the other, oxidation, both to be avoided. Oxidation leads to loss of primary aromas and fruit flavours, with browning of colour and deadening of flavour.

This means a constant balancing act depending on the style of wine being made: at some point towards the reduction end for a pure varietal expression of unoaked sauvignon blanc; near the oxidation end for tawny port and Oloroso sherry. It’s all about using the right amount of oxygen, at the right times to make the desired style, and there are no linear equations that make the process easy.

From the moment grapes are picked, oxidation is a threat

From the moment grapes are picked, oxidation is a threat, but some processes need oxygen. It is essential for the efficient start and smooth completion of alcoholic fermentation. Other benefits of appropriately timed oxygen introductions are more commonly seen in reds: colour stabilisation, flavour intensification, the building up and softening of tannic structure, limiting reduction aromas such as those of onion, rubber, garlic and cabbage, and encouraging natural clarification as less stable phenols are precipitated. It is also thought to help with the amelioration of green characters.

Red wines are thus less susceptible than whites and rosés to oxidation. Red juice and must contains higher levels of phenols (colour, tannin). Oxygenation, the deliberate and wanted introduction of oxygen, usually via air, is important to the resulting quality of red wines. In the cellar, maturing wine in barrel (old if no new flavour is wanted), racking, with smaller or larger amounts of air, and topping up, are ‘oxygenation’ techniques. Leaving the barrel bung at 12 o’clock is a more oxygenating technique than at 2 o’clock, where the bung remains below the wine surface.  Micro-oxygenation is the modern apparition.

microbial instability is an additional risk of too much oxygen 

Apart from direct oxidation, microbial instability is an additional risk of too much oxygen in contact with wine. Wine-spoiling acetic acid bacteria grow when there is enough oxygen. This is why volatile acidity can be a precursor of oxidation, though there are notable reds such as Château Musar and Grange, where otherwise high levels of volatile acidity are integral to the style. Also brettanomyces yeast multiply in must and wine in the presence of oxygen, producing medicinal, horsey, elastoplast aromas, a little of which add complexity, a lot of which is a fault.

At the opposite end of the spectrum, reductive winemaking aims to avoid oxygen all the way, and is mainly used on white wines, especially aromatic and semi-aromatic grape varieties. Oxygen reacts with phenolics, of which whites have fewer than reds, and whites need their few phenolics to contribute to aroma and mouthfeel. Dry ice, inert gases, closed winemaking kit, and regular sulphur dioxide monitoring are the hallmark of reductive winemaking. Low temperature is another tool of reductive winemaking, because reactions are slowed, but oxygen is absorbed more easily at low temperatures.

The issue of ‘reduction under screwcap’ is taken out of the clutches of the closure camps, and the science laid bare.

Stink bombs and the grail of thiol

Oxidation is one thing.  But at the opposite end of the spectrum is ‘reduction’ and wine faults at this end of the continuum have just got trickier.

We’re all familiar with the reductive-state stink-bomb, rotten egg odour of hydrogen sulphide (H2S) and the garlicky, onion and rotten cabbage odours of mercaptans, These are expressions of faulty wines.   

Both of these groups of compounds are sulphur-containing off-odours. As we’ve started to understand more of the chemistry about these and other volatile sulphur compounds (VSCs), some are being implicated as positive flavour attributes in some wines.

Sulphur terminology:

• Sulphides: the sulphides and disulphides. (Though the term is loosely used to mean the off-odour VSCs)  
• Thiols:  A group of compounds previously known as mercaptans. These are susceptible to oxidation.
• H2S: hydrogen sulphide
• VSC or sulphur-like odours (SLOs): Any of the sulphur compounds which can contribute positively or cause reductive off-flavour in wine.

VSCs, and their management, are trendy, because they’re not all bad. Something like a bit of the devil you know does you good. There are some important issues for wine flavour and correctness to be understood:

1. VSCs are related to yeast fermentation, with aroma precursors in grapes
2. Winemaking can clear up some VSCs, but not all
3. There not all bad.  Some VSCs can be good, actively sought-after, complexing components in a wine
4. VSCs are thought to be more frequently seen under closures which have a negligible air ingress  

The Redox Potential

Wine contains sulphides.  The form that these take changes.  And that form is dependent on how much or how little oxygen is around. 

Think of the tightrope walker’s balancing pole.  At one end is ‘reduction’ and at the other is oxidation.

Balance is achievable somewhere along the middle section of the pole. With a bit of oxygen a wine may move a bit towards the oxidation end (high redox potential).  With less and less oxygen, a wines moves towards the reduction end (low redox potential).

Depending on various vineyard and winemaking circumstances, the balance may be tipped a little too much one way or the other, i.e. towards oxidation or reduction.

The lower the redox potential the greater the risk of reduction. The greater the opportunity to form VSCs – good ones and bad ones.

Adrian Coulter, senior oenologist at the AWRI, specialises in off-odour VSCs. He explains the basics part one – chemistry: “One of the ways mercaptans are thought to be formed is by reaction of H2S with wine components. Mercaptans can oxidise to disulphides. Disulphides can be reduced back to mercaptans”.

Dr Alan Limmer, owner of a PhD in chemistry and Stonecroft winery in New Zealand, explains the basics, part two – smell: “Mercaptans oxidise to disulphides. The sensory impact is very different. Disulphides are 20 to 40 times less smelly than mercaptans. So the sulphides have not gone away, with racking for example, just changed to disulphides which smell less. Disulphides can revert back to mercaptans (under reducing conditions). The smell profile changes as soon as you do this. Mercaptans and H2S stink.”

The reversibility of most redox reactions is where some of the problems arise. Non-smelly disulphides can be ‘reduced’ back to very smelly thiols.

Some of the more common VSCs

The good, the bad and the ugly

Generally thiols are both bad and ugly.  H2S and mercaptans definitely stink and cause problems in wine.  But, Limmer says: “The sensory effect of them changes markedly as we add on atoms. For example, figuratively speaking from sewer to lavender with a one-atom change. We get complex flavour compounds which have characters that we associate with being positive. Fresh cut hay for example.”

But there’s no way (yet) to keep the good ones and remove the bad ones.

So much for the bad and the ugly. Some thiols confer attributes we consider positive flavour aromas. Dimethyl sulphide  (DMS) is a case in point.  Coulter said: “At low concentrations, DMS has a blackcurrant aroma and might contribute toward the body of aged white wines. At higher concentrations it is generally regarded as a fault, with cooked corn, cooked tomato and molasses notes.”

A Rhône study into DMS in syrah and grenache showed DMS was released during bottle ageing, giving flavours of ‘truffle, undergrowth, black olive’ usually associated with maturing wines.

But it’s from sauvignon blanc, notably work done by Professor Denis Dubourdieu in Bordeaux, that much new understanding has arisen.  Sauvignon blanc would be nothing without VSCs, plus, of course, a healthy dollop of that defining green pepper note, methoxypyrazine.

VSCs and sauvignon blanc

Some of these VSCs, for example, 4MMP, 3MH and 3MHA, also play a role in the flavour of other grape varieties such as chenin blanc,  gewürztraminer, manseng, muscat and riesling. 

The grail of thiol – control and management

VSCs are in wine. Full stop. Yeast make H2S and other reductive odours. The amount of nitrogen during fermentation is a fundamental driver of VSC development.  Limmer said: “Throughout fermentation wines accumulate a range of sulphide compounds. What varies is the amounts of these compounds. And the pattern varies from wine to wine. If there is insufficient nitrogen, yeast use sulphur-containing amino acids as a nitrogen source. There also appears to be an underlying yeast genetic link to the sulphide profile.”  So nitrogen is just one of the determining components.   

Keeping yeast happily fed with nitrogen keeps down the volume of VSCs.  Aerating wine oxidises mercaptans to disulphides, and ‘blows off’ H2S, which is oxidised to elemental sulphur.  Regularly stirring lees avoids reducing conditions building up.  Leaving lees without stirring builds up reducing conditions.

Copper sulphate fining can remove H2S and, to a certain extent, mercaptans, but will not touch the disulphides. Coulter said: “You have to create reducing conditions to get disulphides back to mercaptans.  Then you can copper fine. If a wine is in barrel, put it in tank, add SO2, wait a week and do a copper trial.”

Additionally, copper sulphate fining doesn’t distinguish between the desirable thiols and the undesirable ones.  It takes them all.

Prevention is always better than cure.  Professor Dr Doris Rauhut, specialist in off flavours in wine at Geisenheim Research Centre said: “Bad thiols mask the good ones. If you can minimise the bad ones, the good ones are more noticeable. Positive thiols occur in precursors (in the grape) which are not volatile. They are released during yeast activity.  

“It’s often better to avoid the development of bad thiols early enough, then you don’t need further treatment in the wine” such as copper fining. This way the overall flavour of some varieties can be optimised. 

Rauhut is working on early detection systems so that with the right yeast strain and optimal initial nutrient composition the fermentation can be controlled early to prevent off flavour VSCs from forming. Leaving the good ones on centre stage.

Genie in the bottle

While aeration and copper sulphate fining are useful tools for bulk wine, once the wine is bottled chemical reactions don’t stop.  They are dependent on the redox potential – how much oxygen is available post bottling, or can ingress post bottling, for the oxidation of smelly thiol back to non-smelly disulphide.  They’re also dependent on the amount of reducing agent – free SO2.

In the opposite vein, if non-smelly disulphides are bottled these are then free to reduce back to smelly mercaptans. (under reducing conditions).  Changes keep going over time, which explains why some wines are clean and fresh, fine and dandy one moment, but 12 to 24 months later, under reducing conditions, smelly thiols have taken over the front seat.

It’s not just the aroma that’s affected. Limmer said: “The biggest effect of VSCs is on the palate. Sulphide components affect the palate more than the aroma. In whites, the palate is shortened, with a clunky, hard, bitter, finish. The wine doesn’t finish in accordance to expectation.

“In good reds there is a silky integration of structural tannin. When sulphides are on top of this, the tannins get hard, sometimes with a green tannin edge. When you take out the sulphur, the velvety texture of tannin comes through again, which is how the winemaker intended.”

Desperately seeking VSCs

A number of Australian producers are looking to gain complexity in their wines by retaining some of these VSCs.  Steve Webber, manager and winemaker at De Bortoli said he’s looking for “minerality, texture and the right kind of mercaptans”.  He said it’s about “extracting the right phenolics from the stems and the skins – sulphur-containing amino acids.”

But get this wrong you lose mineral definition on the palate and you get bad mercaptans. “Good mercaptans” he said “highlight mineral, flinty, gunsmoke characters” in sauvignon blanc.  He also talked of the mealy, mineral characters he’s looking for in chardonnay. 

The six million dollar question – are screwcaps implicated? 

This is emotive and politically charged territory. The arguments currently revolve around air ingress, which is not necessarily a closure-specific aspect, and there is not yet a definitive answer.

Limmer said: “When the wine is bottled the compounds (disulphides) are still in there, you just can’t see them. Over time they break down and revert (reduce) to thiols.  There is an accumulation of smelly thiols in bottle.  And we’re seeing a difference under screwcap and cork. The same problem has been created under cork by sealing the top of the cork. Trials show the biggest problem is in wine sealed under ampoule.  It is suggested to be an oxygen ingress issue, but we don’t know this for certain.”

At the Geisenheim Research Centre in Germany, closure specialist Dr Rainer Jung said: “Our experience is that it hasn’t happened here.  We have used screwcaps since the 1970s for experimental wines and I never had the problem that wine with screwcap has a reductive character.”

Jung speculated why, saying: “In Germany few people put inert gas in the headspace so there is more oxygen in the headspace. In the first year of storage the loss of SO2 is little bit higher for screwcap-bottled wine because of the bigger headspace (in comparison to cork). Except if you add CO2 or nitrogen in the headspace. And it equalises after a time because screwcap is more dense.”

Coulter said: “Winemakers work hard to avoid sulphur off-odours, however, given that we don’t yet understand all the reactions that occur and the factors that influence them, the formation of these compounds is sometimes beyond winemakers’ control.”

In the end

VSCs are the new frontier of winemaking.  We don’t yet know how to keep the good ones and remove the bad ones. We don’t know how much oxygen is just the right amount, but it varies with grape variety, viticulture and winemaking. We don’t know what is a good closure.  And we don’t know how much VSC is a good thing or how to keep it at an optimum level in bottle.  It’s a little akin to the volatile acidity issue or the brett issue, where a little bit can add complexity and too much is a fault. And views on when one moves from good to bad vary with the individual.

Excess sulphur dioxide is not considered to be an off-odour VSC

• SO2 is an antimicrobial and antioxidant.
• SO2 is not a sulphide.
• Enough needs to be added to prevent the wine from oxidising, but not so much as to destroy flavour.
• Too much free SO2 smells sulphury, like burnt match, a sharp, acrid smell.
• However, being a reducing agent SO2 can help break down non-smelly disulphides to smelly thiols.